Abstract:

Aiming at the premature convergence problem which the particle swarm optimization algorithm suffers from, a chaos particle swarm optimization based on adaptive inertia weight is proposed. Firstly, chaotic sequence generated by cube map is used to initiate individual position, which strengthens the diversity of global searching. Secondly, adaptive inertia weight is adopted to improve the convergence rate. Furthermore, chaos perturbation is utilized to avoid the premature convergence. The results of the simulation experiment show that the convergence rate and the precision of the improved algorithm are obviously enhanced, and the algorithm can effectively avoid the premature convergence problem.

Key words:  particle swarm optimization; cube map; adaptive inertia weight; chaos perturbation